Reply to Letter to the Editor: “Chest imaging using signs, symbols, and naturalistic images: a practical guide for radiologists and non-radiologists”

Author: Stefano Palmucci

Affiliation: Radiology Unit 1, Department of Medical Surgical Sciences and Advanced Technologies “GF Ingrassia”, University Hospital “Policlinico-Vittorio Emanuele”, Catania, 95123, Italy.

Dear Editor,

In the recent letter regarding my article [1], Kitami et al pointed out that the terms galaxy sign and cluster sign could be considered misnomers. It was a great pleasure to read their observations and descriptions; thanks for the appreciation regarding my article.

The galaxy sign – to the best of my knowledge – was first reported in literature in 2002, when Nakatsu et al [2] described “large nodules […] surrounded by many tiny satellite nodules. These findings were considered to simulate the appearance of a galaxy”. The sarcoid galaxy was encountered in 16 (27%) of 59 patients of this mentioned study: looking at the figures of this paper, authors provide explanation of sarcoid galaxies – showing “large parenchymal nodules” surrounded by tiny nodules [2].

The sarcoid galaxy sign has also been reported in the fine paper published by Criado et al [3], where they provide same description already outlined by Nakatsu: “Fine nodular opacities are seen around the large nodules […], and small low-attenuation spots that correspond to the spaces between partially coalescent small nodules are visible peripherally.” In another article published by Herraez et al [4], the cluster sign has been described as the presence of “multiple small punctiform nodules in the peripheral regions of the lung in patients with sarcoidosis” [4]. This appearance is again reported in the paper published by Criado [3], which indicates that the sarcoid cluster “consists of multiple micronodules distributed along the lymph vessels on high-resolution CT images.”

These HRCT signs seem to refer to a different degree of profusion of granulomas within the lung: in the first model, we have a large nodule/consolidation and some tiny nodules located in the adjacent parenchyma; in the second model, we do not observe a big nodule – but a constellation of small nodular opacities.

Therefore, according to papers published by Nakatsu in 2002 [2] and then by Herraez in 2009 [9], we have the galaxy sign and the cluster sign; both are related to a granulomatous disease, namely sarcoidosis.
The letter I received is very interesting because it suggests clarification of the difference between “globular cluster sign” and “open cluster sign”: the former is the appearance reproduced by a large nodule/mass with some nodular lesions located around it, whereas the latter is the morphological condition due to a lot of small and punctiform nodules in the lung parenchyma.

Looking at the clarifications reported in the letter, and considering definitions reported on different websites [5], globular cluster stars and open cluster stars represent parts of galaxies: more in detail, open clusters could also represent the spiral arms of the galaxy, or may resemble the appearance of irregular galaxies [6].

Their appearances should be related to different profusions of granulomatous lesions in sarcoidosis, so that “globular cluster and open cluster signs should be used instead of the galaxy and cluster signs” – as suggested in the letter regarding my article. We have a strong correspondence with the appearance of granulomas in sarcoidosis.

In my opinion, however, we cannot reject the first definition reported in literature – the galaxy sign.
As reported in the letter, “The galaxy is disk-shaped and is surrounded by a spherical halo”; and a galaxy consists of globular cluster and open cluster stars, which reproduce respectively the globular cluster sign and the open cluster sign found in sarcoidosis.

However, I’m not an astronomer: even if the associations provided by the authors of the letter are more detailed, I’m not sure that the galaxy appearance in sarcoidosis should be replaced by the terms globular cluster and open cluster – terms which refer to stars and not to galaxies.

More in detail, we need to focus on the image or appearance of a galaxy: it is disk-shaped, and consists of a dense central part – the large nodule found in sarcoidosis – surrounded by a spherical halo (the halo sign depicted on CT images) – where small punctuate stars are also clearly recognizable (the other small tiny nodules adjacent to the central part). So, the term galaxy can still be associated with sarcoidosis: we do not have the resolution to depict which kind of stars are located into the galaxy. And the spiral arms of some galaxies, or some galaxies as reported on different websites, clearly reproduce the appearance of multiple, small, punctuate nodules.

I’m not sure that we could exclude this simple association with a galaxy, to describe the profusion of nodules; in my opinion, we could still use the term “galaxy”. Clearly, for differentiating degree of profusion of nodules within the lung, a univocal correspondence is better obtained describing the open cluster and globular cluster appearances of stars.

Thanking you for the correspondence – associations with signs, symbols and naturalistic images remain a fascinating topic for radiologists.



[1] Chiarenza A, Esposto Ultimo L, Falsaperla D et al (2019) Chest imaging using signs symbols, and naturalistic images: a practical guide for radiologists and non-radiologists. Insights Imaging 10(1):114
[2] Nakatsu M, Hatabu H, Morikawa K et al (2002) Large coalescent parenchymal nodules in pulmonary sarcoidosis: “Sarcoid galaxy” sign. AJR Am J Roentgenol 178:1389-1393
[3] Criado E, Sánchez M, Ramírez J et al (2010) Pulmonary sarcoidosis: typical and atypical manifestations at high-resolution CT with pathologic correlation. Radiographics 30(6):1567-86. doi: 10.1148/rg.306105512.
[4] Herraez OI, Alonso ON, Lopez GL (2009) The “sarcoid cluster sign”. A new sign in high resolution chest CT. Radiologia 51:495-499
[5] Chaisson EJ, Sawyer Hogg-Priestly H, Fernie JD. Star Cluster ASTRONOMY. From ENCYCLOPEDIA BRITANNICA –
[6] Open Cluster, from Wikipedia –